Could the CARE–SHAKTI intervention for injecting drug users be maintaining the low HIV prevalence in Dhaka, Bangladesh? Anna M. Foss 1 , Charlotte H. Watts 1 , Peter Vickerman 1 , Tasnim Azim 2 , Lorna Guinness 1 , Munir Ahmed 3 , Andrea Rodericks 4 & Smarajit Jana 5 HIV Tools Research Group, Health Policy Unit, Department of Public Health and Policy, London School of Hygiene and Tropical Medicine (LSHTM), UK, 1 International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B), 2 Injecting Drug User Team Leader, CARE–Bangladesh, 3 Assistant Country Director—Programs, CARE-Bangladesh, 4 and CARE–India, Delhi, India (previously CARE-Bangladesh) 5 ABSTRACT Aims To explore whether the low HIV prevalence observed in Bangladesh results from prevention activities, this study uses mathematical modelling to estimate the impact of a needle/syringe exchange intervention for injecting drug users (IDUs) in Dhaka, Bangladesh. Design Epidemiological, behavioural and intervention monitoring data were used to parameterize a dynamic mathematical model, and fit it to National HIV Sero-surveillance data among IDUs (2000–02). The model was used to estimate the impact of the intervention on HIV transmission among IDUs and their sexual partners. Setting Dhaka, Bangladesh, where the HIV prevalence has remained low despite high-risk sexual and injecting behaviours, and growing HIV epidemics in neighbouring countries. Findings The model predicts that the intervention may have reduced the incidence of HIV among IDUs by 90% (95% CI 74–94%), resulting in an IDU HIV prevalence of 10% (95% CI 4–19%) after 8 years of intervention activity instead of 42% (95% CI 30–47%) if the intervention had not occurred. Conclusions The analysis highlights the potential for rapid HIV spread among IDUs in Dhaka, and suggests that the intervention may have substantially reduced IDU HIV transmission. However, there is no room for complacency. Sustained and expanded funding for interventions in Dhaka and other regions of Bangladesh are crucial to maintaining the low HIV prevalence. Keywords Bangladesh, harm reduction, HIV prevention, injecting drug users, mathematical model projections Correspondence to: Anna Foss, Department of Public Health & Policy, London School of Hygiene andTropical Medicine, Keppel Street, London WC1E 7HT, UK. E-mail: anna.foss@lshtm.ac.uk Submitted 17 February 2006; initial review completed 16 May 2006; final version accepted 18 July 2006 INTRODUCTION Bangladesh is at a critical stage in its response to the HIV/AIDS epidemic. Despite growing epidemics in neigh- bouring India and South-east Asia [1], the HIV preva- lence in Bangladesh remains very low, even among high- risk groups such as injecting drug users (IDUs) and sex workers [2]. There is ongoing debate about why this is so. Some suggest it could be due partly to the protective effect of male circumcision, and stricter religious and social controls over behaviour in a predominantly Muslim nation that is more insular than India [3]; yet national reports provide evidence of widespread high-risk behav- iours throughout Bangladesh [4,5]. It may be that the early intervention activity of governmental and non-governmental agencies, targeted at groups most vulnerable to HIV, has played an important role in main- taining the low prevalence. Particularly as early in an epidemic HIV often spreads most rapidly among IDUs and is then transmitted through their sexual partner networks to the general population, interventions targeted at IDUs have the potential to play a substantial role in limiting more widespread transmission [6,7]. As part of the CARE–Bangladesh HIV prevention pro- gramme, the CARE–SHAKTI IDU intervention began in Dhaka in 1998. The main intervention activities were needle/syringe exchange, abscess management, treat- ment of sexually transmitted infections (STIs), education and condom distribution [8]. Broader benefits include reducing harassment of IDUs by police and local ‘goons’ RESEARCH REPORT doi:10.1111/j.1360-0443.2006.01637.x © 2006 The Authors. Journal compilation © 2006 Society for the Study of Addiction Addiction, 102, 114–125